I think the author of this piece misses a point. Let me take the example of the American Physical Society (APS). They have membership fees (required to go to their conferences) and journals which charge money for access. As with all academic publishing, you sign a copyright waiver (but you are allowed to post your article on your own homepage--also, for many of us, we initially publish things on arxiv.org in preliminary form (before peer review). So, you have two sources--the author if they choose to distribute it, and the arxiv for many papers (excepting breaking high profile papers that might appear in Science or Nature). Overall, the APS is relatively good--they provide much cheaper access rates to poorer institutions domestically (USA) and internationally.<p>Now, you might wonder--what is the value added that publishers provide? Governments pay for the research, government sponsored researchers do most of the initial typesetting/editing, and government sponsored researchers do the peer review. The answer is that publishers:
1) Screen the papers to send out for peer review
--there is a lot of noise in submissions. Is this paper interesting or not? Does it pass the "sniff" test on technical correctness?
To do this first pass requires editors with domain level knowledge of the field--they can skim the article if they are not experts and guess about impact/sniff/etc. They are familiar with the reputations of researchers. Essentially these are people with at least a PhD in the field and this is not cheap.
2) Again, the editors are acquainted with potential reviewers in the field. They have amassed a database of people who are good reviewers. For example, referee A always critical? Is referee B always a softee? Does referee C respond in a timely fashion? Does referee D always have his reviews overturned on appeal to an additional referee? Etc.<p>So, essentially, what you are paying for is screening out signal to noise. The first thing you might think of is, well, let the government step in. I think this would be very bad. Why? Because it would then be easy for the government to determine what gets published and what doesn't. Think climate change....<p>Next, you might think--ok, well what about open access? Again, the problem is that many open access journals (where the authors pay and the public reads for free) is that the prices to authors steadily increases and unless it is mandated, authors may seek other journals. Especially if the most prestigious journals are not open access....
Finally, you might ask, well, what about a technical solution?
Hacker News for academic publishing, complete with automated tracking of referees, calculations of reputations based on citations, commenting, etc. But, then you'll bump into problems of avoiding winner take all and that for academic publishing, you want the comments and referee process to be deliberate and considered.<p>What I would suggest is that for the computer science community to work harder on posting things to the arxiv before submitting to ACM journals and if that's not possible, then lobbying for journals to accept this. Then, for people who are ok with the rough version, they can access it and for people with need/resources, they can access the final, polished, piece.<p>In the meantime, try writing the author if you need a paper. In the physics community, many authors are happy to send you a pdf file. One time there was a paper from an expensive publication that we didn't have access to that I wrote the author to request a copy from. He actually sent a reprint--from Japan!